Detection properties of indium-111 and IRDye800CW for intraoperative molecular imaging use across tissue phantom models

Abstract

Significance: Intraoperative molecular imaging (IMI) enables the detection and visualization of cancer tissue using targeted radioactive or fluorescent tracers. While IMI research has rapidly expanded, including the recent Food and Drug Administration approval of a targeted fluorophore, the limits of detection have not been well-defined.

Aim: The ability of widely available handheld intraoperative tools (Neoprobe and SPY-PHI) to measure gamma decay and fluorescence intensity from IMI tracers was assessed while varying characteristics of both the signal source and the intervening tissue or gelatin phantoms.

Approach: Gamma decay signal and fluorescence from tracer-bearing tumors (TBTs) and modifiable tumor-like inclusions (TLIs) were measured through increasing thicknesses of porcine tissue and gelatin in custom 3D-printed molds. TBTs buried beneath porcine tissue were used to simulate IMI-guided tumor resection.

Results: Gamma decay from TBTs and TLIs was detected through significantly thicker tissue and gelatin than fluorescence, with at least 5% of the maximum signal observed through up to 5 and 0.5 cm, respectively, depending on the overlying tissue type or gelatin.

Conclusions: We developed novel systems that can be fine-tuned to simulate variable tumor characteristics and tissue environments. These were used to evaluate the detection of fluorescent and gamma signals from IMI tracers and simulate IMI surgery.

Keywords: fluorescence; intraoperative molecular imaging; radioactivity; three-dimensional printing; tissue phantom; tumor detection.

Fig. 1

Signal from TBTs decreases as the thickness of the overlying phantom increases. (a) The gamma decay signal (in cps) of Indium-111 in TBTs decreases as thicker porcine tissue is placed between the TBT and Neoprobe^® (baseline signal is at a depth of 0 cm). (b) The fluorescence signal (in MFI) of IRDye800CW in TBTs decreases with porcine tissue between the TBTs and SPY-PHI camera (baseline fluorescence is at a depth of 0 cm, where there is no intervening tissue). (c) Comparison of the maximum tissue thickness (in mm) through which 5% of the baseline (0 mm) gamma decay from In-111 and fluorescence from IRDye800CW in TBTs can still be detected. (d) Gamma decay signal (in cps) from In-111 and (e) fluorescence (in MFI) from IRDye800CW in TBTs similarly decreases with thicker gelatin phantom between the TBTs and the Neoprobe^® or SPY-PHI. (f) Comparison of the maximum thickness of gelatin phantom through which 5% of the baseline (0 mm) gamma decay from In-111 and fluorescence from IRDye800CW can still be detected.

Fig. 2

Autofluorescence and non-specific fluorescence occur in non-tumor tissues and can be mimicked in a gelatin phantom. (a) Near-infrared autofluorescence (when no tracer has been given) is minimal in all tissues, with MFI less than 1. (b) Non-specific fluorescence (from DTPA[In-111]-antiGD2-IR800 tracer outside of GD2-expressing tumors) occurs and could obscure tracer specifically accumulated in GD2-expressing tumors. (c) IRDye800CW can be added to gelatin phantoms to mimic the non-specific fluorescence observed in non-tumor tissue in vivo.

Fig. 3

Signal from TLIs also decreases as tissue phantom increases in thickness between TLIs and detectors. (a) Gamma decay signal from In-111 in TLIs decreases as additional porcine tissue is placed between the TLI and Neoprobe^®. (b) Fluorescence intensity from IRDye800CW in TLIs also decreases with more intervening porcine tissue. (c) At least 5% of the baseline gamma decay signal (signal measured with no intervening tissue nor gelatin) can be detected through significantly thicker tissue than at least 5% of the uncovered fluorescence signal. (d) Gamma decay from In-111 in TLIs decreases with intervening gelatin phantom in a similar pattern, regardless of the size of the TLI. (e) Fluorescence signal decay through gelatin is not visible through thicker gelatin phantom when TLIs are larger ($p=\mathrm{ns}$). (f) At least 5% of the baseline gamma signal penetrates through more gelatin phantom than the fluorescence signal, regardless of the TLI size.

Fig. 4

(a) Tracer-bearing xenograft tumors (arrows) were embedded on the back side of a section of the porcine body wall, which was then carefully turned over so the skin was on top. (b) For IMI experiments, a blinded investigator used the Neoprobe^® to determine the location of the embedded tumors (arrow pointing to counts of 13 cps). (c) As the investigator dissected closer to the tumor, the gamma decay signal increased (arrow pointing to increased counts of 273 cps). (d) Once nearly exposed, the investigator used the SPY-PHI camera to assess the fluorescence of the tumor (fluorescent image in background), which is then shown to have clear margins when removed. (e) After tumors were resected, surgical defects in the porcine tissue were measured and (f) defects were smaller when IMI was used ($8.82\pm 4.51{\mathrm{cm}}^3$) versus when IMI was not used ($20.82\pm 2.1{\mathrm{cm}}^3$; $p<0.0001$).

Fig. 5

(a) Custom 3D-printed molds with (i) tumor-like inclusions exposed and (ii) under gelatin tissue phantom as well as (iii) TBTs exposed and (iv) under gelatin tissue phantom. (b) TinkerCAD schematic of custom 3D-printed molds including (i) stadium style plate for tumors and TLIs under gelatin, (ii) round style plate for tumors and TLIs under gelatin, (iii) step-style plate for measuring background signal from gelatin with dilutions of IR800, and (iv) a hollow frame (2, 4, 6, 8, or 10 mm tall) for cutting porcine tissues to specific thicknesses.

Fig. 6

(a) Schematic of TBT resection, placement on plates, coverage by porcine tissue, and measurement by Neoprobe^® and SPY-PHI. (b) Uncovered white-light (left) and NIR fluorescent (right) images of TBTs on 3D-printed plates. (c) Sample TBT coverage by each type of porcine tissue.

Fig. 7

Overlay images of fluorescence of IR800 containing tumor-like inclusions, at depths of (a) 1 mm, (b) 6 mm, and (c) 10 mm of gelatin tissue phantom.

Fig. 8

Images of surgical defects for simulated tumor nodule resection with IMI (top) and without IMI (bottom).